Automated AMD Assistance
Could a new image analysis technique contribute to earlier or more accurate diagnosis of AMD?
What do orbiting satellites and ophthalmic imaging techniques have in common? Both may soon benefit from computerized pattern recognition to improve their accuracy. Researchers from the University of New South Wales applied multispectral, unsupervised pattern recognition – the same method currently used to develop satellite maps – to 184 fundus images to see whether or not the technique might improve their quantification and classification, leading to better diagnosis of diseases like AMD (1).
The idea itself is not new – senior author Michael Kalloniatis says the possibility first occurred to him 14 years ago – but the technology has only now reached the point where inspiration can become reality. “We could only test the hypothesis once these imaging techniques became more established,” he said (2). And that day seems to have arrived. In Kalloniatis’ study, which examined 10 normal eyes and 36 with intermediate AMD, the pattern recognition approach demonstrated 74 percent sensitivity and 98 percent specificity in detecting AMD lesions, and further correctly classified 75 percent of large drusen and 68 percent of pigmentary abnormalities.
Does this mean ophthalmologists can now leave the diagnostic work to their computers? Not quite. The method is a powerful new way to integrate multiple imaging modalities and combine their strengths, but it still relies on a supply of accurate and appropriately processed images from each individual modality. Unprocessed images, the authors warn, are susceptible to errors caused by both physical (choroidal vasculature visibility, fundus pigmentation) and technical (non-macular signatures, background intensity gradient) variations.
This approach, like all imaging techniques, is not intended to replace traditional imaging and funduscopy; rather, the authors hope it will serve as an enhancement. And who knows – one day in the not-too-distant future, this technology might even form part of an automated diagnostic or clinical decision support tool that could enable the earlier detection of AMD.
- Ly et al., “Multispectral pattern recognition reveals a diversity of clinical signs in intermediate age-related macular degeneration”, Invest Ophthalmol Vis Sci, 59, 1790–1799 (2018). PMID: 29610844.
- D Smith, “Satellite imaging techniques may help reduce preventable vision loss” (2018). Available at: bit.ly/2KmEnWm. Accessed May 14, 2018.
While obtaining degrees in biology from the University of Alberta and biochemistry from Penn State College of Medicine, I worked as a freelance science and medical writer. I was able to hone my skills in research, presentation and scientific writing by assembling grants and journal articles, speaking at international conferences, and consulting on topics ranging from medical education to comic book science. As much as I’ve enjoyed designing new bacteria and plausible superheroes, though, I’m more pleased than ever to be at Texere, using my writing and editing skills to create great content for a professional audience.
